The overall goal of this grant is to identify clinical predictors of episodes of hospital-based care in people with asthma. This study is of great practical importance because asthma is a common condition (3-5 percent of the population) which has a recent, unexplained rise in morbidity and mortality. Many previous epidemiological studies have focused on predictors of hospitalization for people with asthma seen in the emergency room. These studies are of limited value for the practicing physician who sees the vast majority of patients with asthma. This study will develop a key piece of information needed for outpatient care of patients with asthma; a profile of patients at risk for severe, potentially life-threatening exacerbations of asthma. The experimental design is a prospective study of 800 patients with asthma drawn from a large, pre-paid health care plan. The key advantage of this population is that all the care occurs within the health plan. Characteristics which will be identified include demographic factors, socioeconomic status; patient characteristics such as tobacco use, atopy, pattern of medication use, self-reporting of compliance/adherence, attitudes about asthma; historical assessment of indoor air quality; characteristics of asthma such as duration of asthma, lung function, and variation in asthma symptoms; and speciality of physician. The outcome measures will be obtained over a three year period and will include all episodes of hospital-based care including hospitalizations, emergency room visits, and urgency care clinic visits. Based on these data, models will be developed which will assign relative risk to each of the independent variables. These models will be validated in a subset of the population. Results from this study will for the first time enable a profile to be constructed of the high risk patient with asthma which includes identification of modifiable risk factors. These results can be used for physician and patient education programs as well as to target medical intervention.